This collaborative project will develop and integrate complementary mathematical and experimental microbiology approaches to identify gene targets for the design of antibiotic drug combinations that can select against resistance. Diseases caused by microbial infection are the second leading cause of death worldwide. A major challenge in the treatment of such infections is the development of new classes of antibiotics. Inevitably, the introduction of a new antibiotic into the clinic is accompanied by the emergence of resistant microbes. There are, therefore, two major issues to consider in the development of new antibiotic drugs: 1) the identification of novel targets and 2) the discovery of approaches that will limit the emergence of resistance. The major goal of this project is to develop and apply an innovative methodology that can address these two problems simultaneously. The proposed research builds on the recent discovery that the combination of two antagonistically interacting antibiotics (where one drug suppresses the effect of the other) can select against antibiotic resistance. The applicability of this concept has been limited by the fact that most existing antibiotics do not have a known antagonistic partner. The specific aims are thus to develop approaches that will allow us a) to identify gene targets for the design of antagonistic drugs for existing antibiotics and b) to identify paired gene targets that would permi the development of novel antagonistically-interacting drug pairs. This will be achieved by systematically identifying synthetic rescues, which are genetic interactions whereby the negative phenotypic effect of a gene deletion can be partially compensated by the targeted inactivation of other genes. Such rescues are the genetic counterpart of antagonistically interacting drugs and, as such, present potential targets for the development of novel, antagonistic drug combination therapies that can select against drug resistance. To overcome the combinatorial explosion in the number of possible gene-gene combinations in a bacterium, the discovery of such desired drug targets will be guided by new predictive modeling integrated with directed experimental validation. The project will be focused on Escherichia coli K12, which is a robust experimental system that is amenable to mathematical modeling.